[vsnet-chat 5807] Re: [AAVSO-DIS] survey, pixy etc

[Seiichi Yoshida <comet@aerith.net>]

Dear Taichi Kato,

[vsnet-chat 5797]:
>    I don't recommend to use PIXY for accurate photometry; the software is
> rather for survey purposes.  The most important caveat in using PIXY is
> that the software uses variable aperture (or effective apperture), which
> means the aperture size is different between stars.  The photometry in
> PIXY works something like "SExtractor" i.e. we usually need second-step
> photometry to get accurate results.  The users of older version of PIXY

The PIXY System 2 sums up value of pixels whose value is larger than
the (automatically determined) threshold. It does not do PSF fitting
at all.

But in the MISAO Project, we measure the magnitude of many known
variable stars from wide field images in this method, and report to
VSNET. They may be somewhat inaccurate, especially comparing to the
professionals' results, but somewhat useful for some purposes.

PSF fitting source codes are welcome to be plugged into the PIXY
System 2.

> photometry to get accurate results.  The users of older version of PIXY
> also needs to be aware of the variable coefficient (k) in deducing
> mag = k log(count) + c.

It is the very old version, before 2000 or so.

[vsnet-chat 5803]:
>    Although I am not familiar with other software, I think I know better
> about PIXY implementation.  PIXY is aimed for discovery of new objects,
> but not always for precise measurement.  For this purpose, Yoshida-san
> took a heuristic approach rather than an analytical one.  Although the
> results will be acceptable in many cases, some of the results are expected
> to differ from more analytical results.  If one wish to make scientific
> grade photometry and avoid software-borne bias (e.g. for time-series
> photometry of variable stars), one should better stick to more analytical
> apporoaches.  No software is complete, but the reason of incompleteness
> differs from one to another.

Thank you for your comments in details! Yes, Taichi Kato knows very
well on the PIXY System 2 algorithms, because it had been developped
in discussion at VSOLJ mailing lists communicating with Taichi and
some other members.

But one point, the PIXY System 2 does not concentrate on discovery of
new objects only right now. On the contrary, the main feature is not
discovery of new objects, but to operate a semi-automated project with
huge amount of images, like the MISAO Project. 

One of the main features of the system is capability of matching
between star catalogs and various sorts of images, narrow field to
wide field, with only several stars to with over a dozen of thousands
stars, etc.

There are also many functions and graphical user interfaces to deal
with or search among many images, and to measure and report many
variable stars.

Execuse me, that is off topic. Taichi's comments are correct anyway.

Best regards,

--
Seiichi Yoshida
comet@aerith.net
http://vsnet.aerith.net/

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp